Document Type: Framework
Status: Draft
Authority: Experimentation Brain
Applies To: paid media experimentation design
Parent: Experimentation Brain
Version: v1.0
Last Reviewed: 2026-04-15
Purpose
Defines how hypotheses are constructed for paid media experiments.
Experiments must produce learning, not just performance attempts.
Scope
Applies to:
creative testing hypotheses
message testing hypotheses
audience testing hypotheses
offer testing hypotheses
Core Principle
A test must attempt to answer a clear question.
Unclear questions produce unusable results.
Hypothesis Structure
Each hypothesis must include:
assumption
variable being tested
expected behavioural change
measurement metric
success threshold
Example Hypothesis Structure
Assumption:
problem-focused hooks increase attention rate
Variable:
hook framing
Expected Behavioural Change:
higher click-through rate
Measurement Metric:
CTR
Success Threshold:
15 percent improvement vs control
Hypothesis Categories
message hypothesis
audience hypothesis
creative hypothesis
offer hypothesis
landing hypothesis
Learning Loop Integration
Hypothesis → Test → Result → Interpretation → New hypothesis
Learning compounds when hypotheses build on prior results.
Relationship to Other Frameworks
Supports:
Experimentation Brain Learning Loop Integrity
Experimentation Brain Evidence Hierarchy
Ads Brain Creative Testing Structure Framework
Drift Protection
The system must prevent:
tests without learning objective
vague test assumptions
uninterpretable results
repeated random testing
Architectural Intent
Hypothesis discipline transforms advertising activity into structured experimentation.
Structured experimentation produces compounding learning advantage.